Direct power control of a doubly fed induction generator by using a neuro-fuzzy controller

In this article, a direct power control strategy for a doubly fed induction generation is presented. The control variables - d-axis and q-axis rotor voltages - are determined through a control system that consists of a neuro-fuzzy inference system and a Takagi-Sugeno fuzzy logic controller of first order. The input variables of the control system are the rotor speed, active and reactive power references, and their respective errors. Constant converter switching frequency is achieved, what eases the design of ac harmonic filter and power loss reduction of power converter. Another features are the rotor current-sensorless operation, lower cost implementation and the gain of reliability. To validate the strategy simulation and experimental results are presented for abrupt and linear variations of active and reactive power.